By Neil O’Hara
When the U.S. government first announced its plan to partner with the private sector to help banks sell their underwater mortgage-backed securities, most hedge funds shied away. Bridgewater Associates, in one notable example, said there wasn’t enough leverage from the government and too much in the assets.
But those with more faith in the government’s plan have been richly rewarded for the trade—at least so far.
Six of the eight PPIP (Public-Private Investment Program) managers delivered returns in excess of 34% from inception in late 2009 through September 30, a whole order of magnitude greater than the returns on 10-year Treasurys (+2.6%) and five times the return on credit-related hedge funds over the same period.
The PPIP program was designed to revive the market in private-label residential and commercial mortgage-backed securities, among the assets hit hardest during the financial crisis. For every $1 the selected asset managers raised, Treasury put up $1 of matching equity and $2 of debt, creating $29.4 billion in total buying power to be invested in these legacy assets.
Angelo, Gordon & Co. manages the top performing PPIP fund, AG GECC PPIF Master Fund, which gained 52% during the period. Andrew Solomon, who manages the commercial mortgage-backed securities portfolio, says the firm’s expertise in both residential mortgage-backed securities and the commercial ones enables it to shift capital between the two sectors and among different parts of the capital structure to take advantage of the best available opportunities.
Solomon won’t say which segments have delivered the best returns, but he notes that super-senior triple-A CMBS tranches now trade too rich. These assets yield about 5%, and PPIP’s 1-to-1 leverage can’t boost the return on equity high enough. “Even if you let the borrowing cost float, you get a net return in the high single digits,” he says. “That part of the CMBS capital does not work well for PPIP.”
A year ago, Solomon says, market participants expected cumulative losses in certain CMBS pools to exceed 20%—enough to wipe out the most junior tranche entirely. Those dire forecasts have receded; today the market may expect 15% cumulative losses in the same pools. A bond that was trading in the 20s a year ago is now expected to give investors half their money back plus several years of interest—and likely trades in the 50s.
Although those dodgy bonds have soared as their prospects have improved, some prefer to focus on less risky paper. Jonathan Lieberman, who runs the RMBS portfolio at Angelo, Gordon, says he has focused on bonds that will outperform if the tepid economic environment persists for several years. He favors higher-quality credits, which are less likely to be impaired if foreclosure proceedings are delayed.
“The biggest driver when we acquired assets was the liquidation timeline and the potential for that to extend over the next couple of years,” says Lieberman. “We will get more interest payments if the borrower performs for a longer period, and we are not as dependent on liquidation proceeds for our return.”
Angelo Gordon dug deep into the minutiae of the mortgage pools, using credit data it purchased from credit bureau TransUnion, a level of due diligence at least as thorough as that of underwriters at the time of issuance. Other PPIP managers conducted a similarly detailed review to separate the wheat from the chaff.
The analysis prompted Jeffrey Phlegar, manager of the AllianceBernstein Legacy Securities Master Fund (which was up 40.8% through September), to amend his original plan to split capital equally between CMBSs and RMBSs. Although he acknowledges that CMBS credit spreads have tightened between 1,600 and 2,400 basis points depending on the tranche, investors are ignoring underlying credit problems. The delinquency rate on loans in CMBS pools is 14%, for example and loan modifications on cash-flow-positive properties merely defers the day of reckoning.
AllianceBernstein has put money into all four RMBS sectors—prime, Alt-A, subprime and option ARM—but Phlegar’s portfolio skews toward prime and Alt-A loans. Prime borrowers’ homes are typically worth more than the outstanding loan balance, which makes prime RMBSs a bet on the frequency of default rather than the severity of the loss once default occurs. It’s the reverse for Alt-A; the borrowers have little or negative equity, so expected loss severity drives prices for Alt-A–oriented RMBSs.
Phlegar notes that an RMBS portfolio balanced between frequency and severity plays is relatively insensitive to movements in interest rates at the moment. If the economy remains sluggish, the Fed will keep rates low, which means cheap mortgages will boost housing affordability and reduce default frequency. The Fed will push rates higher only if the economy is stronger, when house prices should stabilize or begin to recover and loss severity will decline. “It reduces some of the near-term interest-rate risk,” says Phlegar. “If the economy recovers, the play will shift to severity.” In practical terms, that shifts the emphasis from prime to Alt-A RMBSs.
Unlike CMBSs, which usually contain loans from all parts of the country, RMBS pools may be more concentrated in particular regions. Phlegar has a bias based on legal considerations toward states where the foreclosure process does not involve the courts. In diversified RMBS pools, he can’t avoid every judicial state, but some have many more undesirable loans than others. It’s one reason Phlegar deems California “one of the more attractive places to have exposure to RMBSs.” The local delinquency rate is about 15% and nonjudicial foreclosure is quick, whereas in Florida 25% of mortgages are either delinquent or grinding their way through foreclosure in the state’s clogged courts. “There is massive oversupply already in the pipe in Florida and more to come,” he says. “Home prices have already fallen by more than 50%, but it is hard to call a bottom or even look for stability.”
Cash flow characteristics drive security selection at the Invesco Legacy Securities Master Fund (+34.4%), run by a team that includes Richard King, president and CEO of Invesco Mortgage Capital. The managers are happy to take capital appreciation when it happens, but the core decision to invest rests on their projections of future prepayments, default rates and loss severity. “We can project cash flows, but we may or may not get price appreciation,” says King.
Invesco’s portfolio is concentrated in RMBSs, especially the top tiers of prime and Alt-A bonds. “On a percentage basis, subprime RMBSs have appreciated more than prime,” says King, “but if you include prepayments at par and a robust coupon on bonds bought at 50 to 70 cents, you end up with better performance on the prime paper, especially on a risk-adjusted basis.”
To some extent, the remarkable performance of PPIP funds reflects a recovery in an oversold asset class. Several triple-A ABX index tranches used as market benchmarks have risen 25% or more this year; adding PPIP leverage would push the return over 40%, a level only four of the PPIP managers beat. The RMBS world isn’t that simple, however; other triple-A ABX indices have delivered returns only in single digits, and one is down for the year.
Market moves more than manager skill account for the returns on RMBSs from fall 2009 to June 2010, reckons Andrew Rabinowitz, chief operating officer at Marathon Asset Management, a New York hedge fund that specializes in distressed debt. But prices have been volatile since then, which has played to one of Marathon’s strengths: an active trading style. It has turned over the PPIP portfolio three or four times already.
“We actively manage the RMBS and CMBS portfolios to find the most attractive asymmetrical risk reward within the capital structure,” says Rabinowitz. “To generate alpha over the past few months has required a keen focus on security selection.” The Marathon Legacy Securities Public-Private Investment Partnership was up 44.3% through September, the second-best result among PPIP managers. Rabinowitz expects greater dispersion of returns among PPIP managers now that the RMBS market is no longer a one-way bet.
So many factors affect the performance of RMBS pools that prices are bound to diverge. The securities serviced by Countrywide Credit have experienced higher than average losses, for example.
While the extraordinary returns PPIP managers earned in 2010 likely won’t likely repeat themselves , Rabinowitz still sees opportunities to capture further price appreciation from careful security selection. “If an RMBS has gone from the 40s to the 60s or 70s, there is still potential for appreciation. The key will be how well you did your security analysis and cash flow analytics to get an exit in the 80s or higher,” he says. AR
